Understanding the new strain of coronavirus: how scientists learn without experimentation

"Be prepared for our understanding to evolve over time. It’s how science actually works," writes NCSE Executive Director Ann Reid about the recent discovery of a new variant of the SARS-CoV-2 virus.

Lattice analogy of the deformation of spacetime caused by a planetary mass.

Check out our entire series explaining the science involved in the coronavirus pandemic. Sign up to receive our coronavirus update each week.

So this new variant of the SARS-CoV-2 virus — is it something to worry about? Is it really more contagious? You may have noticed that news about it always seems to be framed as: “on the one hand” (the new variant may be much more contagious!) and “on the other hand” (we don’t really know whether it’s more contagious!). You may read these different explanations and come to the conclusion that scientists or journalists (or both) don’t really know what’s going on and are just expressing opinions based on their general attitude towards the pandemic. Now, I’m not going to tell you that scientists are immune to the temptation to interpret new data in light of their previous understanding. (In fact, if you think about it, they have to! What else are they going to do, start over with “Hey, what happens when I drop an apple?” whenever they encounter new information?) But in this case, the fact is that nobody really knows yet exactly what to make of this new variant.

You may find the uncertainty frustrating. I do, too. But at least it offers a golden opportunity to gain some appreciation for how scientists learn about phenomena that resist direct experimentation.

One of the more stubborn misconceptions about science — reinforced, I would argue, by school laboratory experiments that always turn out as expected — is that laboratory experiments are the only truly scientific way to learn how the world works. The world didn’t get that memo, and it turns out that relatively few questions can be tested directly by experiment in a laboratory.

But just because we can’t do an experiment in a lab doesn’t mean we can’t learn anything.

But just because we can’t do an experiment in a lab doesn’t mean we can’t learn anything. Sometimes scientists theorize — in other words, make an informed prediction of what might be happening. This is what Albert Einstein did when he predicted that light waves could be bent by gravity — his theory of general relativity. Once his prediction had been made, physicists (including Einstein himself) went to work figuring out how to test it.

General relativity can’t be tested in a lab (or at least it couldn’t be tested in the labs of the 1910s and 1920s). But it could be tested by observation: specifically, by looking at where distant stars appeared to be during a solar eclipse compared to where they appeared to be when the sun was not in the way. Einstein’s theory was confirmed. You might say “So what, who cares about this weird quality of light?” But if you appreciate the GPS capabilities of your phone, you should know that factoring in general relativity makes the GPS location of where you are a little more accurate.

I could sing the praises of other scientists who made predictions based on theory that were later borne out by direct observation or even by laboratory experimentation (I’m looking at you, Charles Darwin!). But theoretical prediction is not the only alternative to direct experimentation. Another critically important way to learn about what’s going on in the world is modeling. There is an outstanding example of how model building contributed to demonstrating that chlorofluorocarbons were destroying the Earth’s ozone layer at the University of California Museum of Paleontology’s Understanding Science website. Additionally, check out NCSE's classroom resources for our activity on Mystery Cubes that give students a taste of what it means to use a model to test their ideas.

It turns out that the most powerful way to figure out exactly what we’re dealing with here is modeling.

Okay, back to the new coronavirus variant. You might think it would be pretty easy to determine whether one variant is more contagious than another. But if you think about it for a minute, you’ll quickly realize that there aren’t any direct approaches available. You can’t separate 1,000 people into two buildings, introduce one variant into each building, and wait to see how many people get sick. (Well, you can if you’re a supervillain. But don’t be a supervillain.) You can test the two variants on cell cultures in the laboratory and see whether one of them invades more cells, but that’s not a direct indication of how the virus would spread in functioning human respiratory systems. So it turns out that the most powerful way to figure out exactly what we’re dealing with here is modeling.

Let’s start with a mental model of the problem. Here’s my best attempt:

Picture a shop full of fine china arranged upon rows of shelves. The shelves stretch from wall to wall and from floor to ceiling, on all four walls of the shop.

There are a few holes in the walls of the shop. Suddenly, through one of the holes comes flying a yellow tennis ball. It strikes a piece of china, which bursts into pieces. The instant it breaks, another tennis ball flies into the room. This one bounces harmlessly off a shelf. All is quiet for a moment. Then another ball bounces in. It strikes a piece of china. Another ball flies in; another piece of china shatters. Soon there are tennis balls flying through the holes every few seconds. Broken crockery is everywhere.

Now imagine that in addition to the yellow tennis balls that have been flying around the room for months, an orange tennis ball comes flying in. When the orange ball hits a piece of china, another orange ball flies into the room. Are orange balls better at hitting things? Before long, there are mostly orange balls flying through the shop. It seems as if the new ball really is more dangerous. But is it?

Maybe; maybe not. Okay, stick with me here for one more exercise in imagination. Imagine there’s another shop full of china next door, and suddenly a hole opens between the shops. One of the orange tennis balls is the first to fly through the new hole. It hits a piece of china. Another orange ball is then released. Eventually a yellow ball gets through, but by then the shop is full of orange balls. Does that mean orange balls are better at hitting things? Not necessarily – maybe they’re dominating because they got into the new shop first.

This is the situation we’re in with the new coronavirus variant that was first detected in England back in September 2020. In some parts of the country, it was soon causing the majority of cases, overtaking variants that had been circulating for months. This probably means that it does spread more easily, but it’s impossible to rule out completely the possibility that the new variant happened to be around in a particular geographic area when restrictions were loosened and the conditions were ideal for its spread.

And that’s where modeling comes in. You can write mathematical equations that capture the variables involved in spread: What percentage of people are wearing masks? How many close contacts, on average, do people in the community have? How many cases are emerging each day, and how many are of each variant? The equations can all be linked together and evaluated by a computer, with as many variations to the equations as you want to try. One of the factors you might vary is how infectious each variant is. Then you can compare the results of the model to what you’re seeing in the community. The equations can be fine-tuned over time.

A recent study from the United Kingdom used just such an approach to model the infectiousness of the new variant (called, catchily, VOC 202012/01). On the basis of their model, the authors estimated that it is 56% more transmissible. Just one week after the release of the original study, the authors released an update with analysis of the spread of the new variant in multiple regions in the U.K. and concluded that the new information made it less likely that differences in transmissibility in the first community studied were due to a founder effect (like the orange tennis ball reaching the china shop first).

By the way, the only reason we’re even aware of this new variant is because the U.K. had put in place a major research effort aimed at extensive sequencing of local coronaviruses. Because the variant doesn’t appear to cause different symptoms or more severe disease, without random sequencing, its spread would have been undetectable against the background of a substantial outbreak already underway in the U.K. So, hats off to as much research as possible.

Unfortunately, recognition of the danger of the new variant came too late for any concerted effort to contain it in the U.K. It has now been identified in 32 countries, including the U.S. (in three different states), in patients with no history of travel to the U.K. or even contact with travelers from the U.K. It’s safe to say that it is already seeded throughout the U.S. and the world. As the authors of the study cited above note, its increased transmissibility may make even tighter restrictions necessary to “flatten the curve” of current outbreaks.

Although the evidence for increased transmissibility so far is strong, the jury is still out on this new variant. The U.K. study has not yet been peer reviewed, and further investigation of how the variant spreads in all the new communities in which it has been found may or may not confirm these initial predictions.

The good news is that the variant is not more lethal, and current vaccines should provide immunity against it. But be on the lookout for more studies using modeling. And — as always — be prepared for our understanding to evolve over time. It’s how science actually works.

NCSE Executive Director Ann Reid
Short Bio

Ann Reid is a former Executive Director of NCSE.

reid@ncse.ngo